developed the disease within the normal incubation limits— three to five days. Subsequently Guiteras and others, on repeating this experiment, have obtained similar results.
The germ not transferable by recently injected mosquitoes.— As might have been partly expected from the blood inoculation experiments already described, it was found that it was only mosquitoes that had fed during the first three days of the fever that were infective. It was further demonstrated that the germ must undergo in the mosquito some necessary developmental process, for it was not until twelve days had elapsed after feeding on yellow-fever blood that the experimental insects could convey infection. Repeated trials made with mosquitoes two to ten days after they had so fed invariably gave negative results; whereas the same insects rarely failed to infect when set to bite non-immunes at any time subsequent to the twelfth day after their yellow-fever blood meal. It was also proved that they retained their infective power for at least fifty-seven days.
These experiments fully explain— (1) The impunity with which a yellow -fever patient can be visited by a non-immune if outside the endemic area; the mosquitoes in the vicinity are not infective. (2) The danger of visiting the endemic area, especially at night; the mosquitoes there are infective and active. (3) The discrepancy between the incubation period, three to five days, of the disease, and the incubation period, fourteen days and over, of an epidemic; the necessary evolution of the germ in the mosquitoes infected by the original introducing patient demanding the space of time indicated by the difference between these two periods. (4) The clinging of yellow-fever infection to ships, buildings, and localities; the persistence of the germ in infected mosquitoes (Stegomyia calopus) which are known to be capable of surviving for five months and probably longer, after feeding on blood. (5) The high atmospheric temperature required for the epidemic extension of yellow fever; such temperature favours the
